A Novel Approach for Hybrid Wavelet Fourier- Hmm Speaker Recognition Voice Verification System Using Wavelets

The paper presents successful experiments on combining two speaker recognition methods into a hybrid system. The rest branch of recognition is an innovative approach based on discrete wavelet-transform. The second one is classic, based on HTK and classication into voice and unvoiced segments. The hybrid solution outperforms both on a small test set. Further tests of this solution will be conducted and reported.. Voice is compared four out of five individual are verified and the result show higher percentage of accuracy. The accuracy of the system can be increased by using a advance pattern recognition The latest technology improvements in such elds as banking, communications and networking require the latest advances in security systems. In the last years a new kind of biometric identication has risen among the others as it is printed in subjects' bodies, it is impossible to lose, almost impossible to duplicate and univocally designates a person. The biometric keys exploit some human characteristics that are dierent and unique in each person such as fingerprints, DNA chains, face shape and voice. Technique such as Hidden Markov Model (HMM).